1. Field of the Invention
The invention relates to sensors applicable for an air conditioning apparatus such as a cooler or heater and, more particularly to a temperature comfort sensing device suitable for obtaining a comfortable air condition by detecting a temperature sensitivity that a typical human being feels on the average.
2. Description of the Prior Art
Generally, a comfortable air condition is dependent upon factors such as air temperature, humidity, radiation rate, air flow, dust (a contamination level), smell, and cleanliness.
For obtaining the comfortable air condition, it is, therefore, necessary to develop sensors capable of detecting human conditions and environments surrounding human beings, and to develop operating mechanisms for carrying out control methods for processing signals from the sensors and achieving various operations.
Recently, an attempt for obtaining a comfortable air condition has been made in air conditioning appliances, in particular, air conditioners, using a predicted mean vote (PMV) value analogized by virtue of the development of neural networks and fuzzy controls, instead of conventional simple temperature controls.
A PMV value is the value which quantitatively expresses as a scale of language the sense of temperature that a person feels and is represented as a function of temperature, radiation rate (wall temperature), humidity, air flow, dress amount and, metabolic rate.
Table 1 shows examples of PMV values. As apparent from Table 1, human beings feel comfortable at a PMV value range of -0.5 to +0.5.
TABLE 1 ______________________________________ Value of PMV Temperature Sensation ______________________________________ +1 Cold +2 Cool +1 Slightly cool 0 Neutral -1 Slightly warm -2 Warm -3 Hot ______________________________________ PMV = f(temperature, radiation rate, humidity, air flow, dress amount, metabolic rate)
FIG. 1 is a block diagram illustrating a temperature control running process of an air conditioner performed by a conventional neural network.
Referring to FIG. 1, PMV variables are deduced from the simulation of an air conditioning environment by measuring basic physical amounts such as intake temperature, variation in intake temperature, ambient temperature, amount of wind, predetermined temperature and direction of wind and inferring dress amount, and metabolic rate. There are a number of difficulties in deducing the PMV variables because it is complicated to interpret the flow of a fluid (an air) and the temperature of an air conditioning environment and because tests for many cases are required.
Also, needed are various sensors such as temperature sensors to measure intake temperature, exhaust temperature and ambient temperature, current sensors to get exhaust amount and hall devices to sense RPM of a motor because much information is demanded for the temperature control by the neural network.
In addition, because the PMV values are indirectly calculated in spite of using so many sensors, a quantity of simulation programs are needed, and an error rate is high in an arithmetic operation based on PMV calculation. There may, for example, be a defect to show many deviations, in that an inference of the radiation rate is based on ambiguous data so that it may vary depending on the assumption about or the state of positioning an air conditioner.